As you point out, the color rendering index of the LED is a function of the phosphors. There are now a number of 'white' LEDs which have better CRI than projector bulbs. It really is only a matter of time before projector bulbs are gone.
Hmmm, I did write "better" and after looking at my CREE bulb specs I realize that 93 is not '99' :-) so they are still a bit worse. In my opinion they are good enough to be equivalent but concede the point that spec wise they are inferior.
On a vaguely related note, any recommendations for white-light (not yellowish-light) LED bulbs for home use that compare favorably with a Reveal incandescent bulb?
In spite of their marketing name they are just bluer lights (2850K versus 2700K for 'warm' and 3000K for 'warm white' so midway between those)
GE has some equivalent LED bulbs that they market under the 'Reveal' moniker to indicate they have a 2850K color temperature. A quick search suggests that Home Depot sells them in the US.
Research or products? I've got one of the CREE [1] LEDs and its pretty good. (my application is building a light tent for macro photography vs a projector though) Philips has some high CRI lights as well which are a bit earlier than the CREE one. Most of the research these days seems focused on more lumens per watt, once the CRI got above 90 most people moved on to the efficiency question.
Thanks!! The idea was to ask about products but links to research would equally well appreciated. Didn't even think to hope for such of goodness :)
IMO each point of CRI really does count, even in the high 90s. I'm using EIKO Solux halogens that have a CRI of 99 as I recall. Excellent light quality, colors are so much better and everything just feels so much more right.
Personally I think CRI is a quite flawed metric, since an incandescent bulb is judged to be 'perfect' and everything else is essentially judged on how close it can get to that, based on a somewhat arbitrary set of measurements.
But I’ll agree with you that LEDs (and especially CFLs) have terrible color rendering, and make pretty unpleasant room lighting.
ChuckMcM: a light source for taking photos is a very different kind of application than room lighting (for which I really dislike all of the LEDs and CFLs I’ve tried, even expensive fancy ones) and also a very different application from projector backlights: projector backlighting is an application where what matters is the interaction between the source spectrum and the transmission spectra of the RGB color filters; the CRI is an almost totally irrelevant metric for projector backlights IMO, except insofar as an LCD was designed to be used with an incandescent backlight (which I doubt a laptop display panel would be); regardless of backlight source, it should be possible with the right equipment (spectrophotometer) to measure the colors projected on the wall and properly characterize the system, and then generate an ICC profile and get pretty accurate color out.
Actually it seems that all laptops now use white LED backlights in their "intended" configuration. That suggests to me that if the phosphors on your projection light source LEDs were equivalent to the ones on the backlight LEDs you would get the same colors coming of the screen as you would from it on a laptop. At that point it becomes a question of the reflection surface as to how well it reflects the spectra you throw at it.
And I agree with you that taking pictures is a different application than room lighting, but there is significant overlap. Essentially I'd like to have a "perfect" broad spectrum lamp for my pictures, and I'm happy to bias the spectrum one way or the other for room lighting. That said, the combination of a broad spectrum room light and a digital photo processing stream lets me get back to the 'original' colors. That was not always true with LEDs but with the high CRI ones it works well for me.
